Dear all,
Now, I have a ‘lat.in’ file as follow (A, B are two transition metal elements, O is interstitial oxygen):
2.940004 0.000000 0.000000
-1.470002 2.546118 0.000000
0.000000 0.000000 4.800997
1.000000 0.000000 0.000000
0.000000 1.000000 0.000000
0.000000 0.000000 1.000000
0.000000 0.000000 0.000000 A,B
0.333333 0.666667 0.500000 A,B
0.666667 0.333333 0.750000 Vac,O
0.666667 0.333333 0.250000 Vac,O
If I do not specify the ‘ref_energy.in’ file and run ‘mmaps’, the ‘atoms.out’ and ‘ref_energy.out’ files contain:
B -6.47738
O -11.3158
A -6.33684
Vac -1.49847
It looks very odd that the energy of Vac is -1.49847 not zero, do I need to set it to zero to get an accurate Ef? By the way, if I want to specify my own ‘ref_energy.in’ file, the energy of O should be the total energy of O2 molecule division by 2 (i.e. Etot(O2)/2). And for A or B, the energy in ‘ref_energy.in’ should be the total energy of A or B division by the number of the atoms in the calculation cell (Etot(A)/N(A atoms)). Is this a correct understanding to get accurate Ef of enumerated structures?
In addition, if I run the ‘mmaps’ without setting ‘nbclusters.in’ file to restrict the number of clusters for pairs, triplets and quaternions, it looks taking much more time to create structures, and some larger CV scores like e+37 existing in the standard output file. If I use ‘nbclusters.in’ file, how many pairs, triplets and quaternions should I set in this file? Is there any basis?
Thanks a lot.